AMD K8L ES coming in December

now we have a real photo of K8L chip, produced in August
http://img90.imageshack.us/img90/2965/k8lps7.jpg

compare it to schematic K8L
http://img246.imageshack.us/img246/4097/amd2im5.jpg

ahhhh :banana::banana::banana::banana:.... :-D

2.5 months to go till the return of the king.

Opteron Deerhound will be announced from April to June 2007, little later (but maybe simultanuosly) Athlon FX and Athlon 64 X4

Whoa !

X4 sounds sweet :D

the first glimpse of my next upgrade.... :slobber:

I wish you a Marry Christmas (XMAS), I wish you a Marry Christmas and a happy new year! :comp10: ;)

Quote:

---

Originally Posted by SparkyJJO

the first glimpse of my next upgrade.... :slobber:

---

Mine too most likely :D

I wanna benches, benches, benches, benches :D


...and overclocking :D

before Deerhound and its derivatives u'll be able to try 65nm K8 Brisbane - typical OC 3,5-3,6GHz on 80-dollars-MB :)

Quote:

---

Originally Posted by MAS

before Deerhound and its derivatives u'll be able to try 65nm K8 Brisbane - typical OC 3,5-3,6GHz on 80-dollars-MB :)

---

How many did you try already to come to that conclusion?

LowRun, do not worry, only one time bugaga

Quote:

---

Originally Posted by MAS

LowRun, do not worry, only one time bugaga

---

Sorry, i couldn't make sense out of this one.

i guess he said that he tried one chip, im not sure though...

Quote:

---

i guess he said that he tried one chip

---

if it were in reality...
Unfortunately K8 Brisbane will be available only in December

Quote:

---

Originally Posted by MAS

if it were in reality...

---

So... you tested it in your dreams? :confused:

I think i just soiled my pants :). i wonder how many pennies i will have to save from now till the day k8l comes out and have a decent price. on a serious note, i hope amd improves on the manufacturing process. i would hate to have 1 out of 4 cores going bad. nothing is more disheartening than having a bad core :). amd should market the k8l as the return of the king similar to lotr: return of the king :). geeks all over the world will go out and buy one :).

Speaking about that, I wonder since K8L will have independant controll over each one of the cores, I wonder if we will see a X3 .... X3 would be 3 processors. One core bad.

~Mike

Quote:

---

Originally Posted by dng29

, i hope amd improves on the manufacturing process. i would hate to have 1 out of 4 cores going bad. nothing is more disheartening than having a bad core :)

---

What makes you think that a core will go bad? (I mean after validation)
This is not the Cell. :p:

Quote:

---

Originally Posted by Piotrsama

What makes you think that a core will go bad? (I mean after validation)
This is not the Cell. :p:

---

i think he means having one core that clocks worse than the others :D

i hope it will not be to expensive

Quote:

---

Originally Posted by galimim

i hope it will not be to expensive

---

I assume we'll have the usual productline going from $400 to $1000 or something.

well if this is more than kentsfield Q6600 im staying with intel ... which is potentially $500 ~ $600 USD

Quote:

---

Originally Posted by MAS

before Deerhound and its derivatives u'll be able to try 65nm K8 Brisbane - typical OC 3,5-3,6GHz on 80-dollars-MB :)

---

I wish personally if I am still on AMD when 65nm comes along I will be happy seeing 3.2ghz

My 130nm 2800+ would do 2.75ghz.

My 90nm AM2 3000+ if I beg and pleed for months might do 3ghz

The WR for 130nm A64 is 4ghz

The WR for 90nm A64 is around 4.2ghz

I really don't think that unless AMD really gets there poop in a group we are going to see 3.5-3.6ghz outside of one of those occasional really great chips.

Wonder if XIPs will be able to show their results? They weren't allowed to with AM2, who knows with K8L. I just hope so.

AMD usually hide their new chips until announce date

one more pict.
http://www.overclockers.ru/images/ne...17/quad_01.jpg

Kent will have 280 square mm chip surface, K8L - only 150 sq. mm

Quote:

---

Originally Posted by MAS

one more pict.
Kent will have 280 square mm chip surface, K8L - only 150 sq. mm

---

That was an old estimate. A more recent estimate from the same person is 283 mm^2.

http://chip-architect.com/news/Quad_vs_Dual3.jpg

thus Kent and Deerhound have equal chip square?

maybe sooner then we think ? remember how long ago that the K8L was suppose to be taped out ? well according to this , it takes 12 weeks from start to finish for a amd wafer , , i think that my math must be failing

http://www.theinquirer.net/default.aspx?article=34781

How AMD bakes its 65 nano Barcelona cakes

Basically, AMD can automate the testing and incorporate the feedback on the fly. When you decide to do 10 per cent more on step 37 of 91, three weeks into a 12 week process, getting accurate and timely feedback is essential. What APM does is allow AMD to pick a set of wafers and apply the special sauce, and track them at every point from then on while cataloging all the metrology details. If it works, and has good yields, it can be made part of the permanent mix very quickly.

either my math is getting bad , or my memory is getting bad

They have to do more testing and build volume for the launch, I doubt we'll see em' earlier than 2Q 2007. Die size seems a little large but doable for AMD, hopfully they can get it down to something a bit more reasonable to improve thier margins.

Quote:

---

Originally Posted by mesyn191

They have to do more testing and build volume for the launch, I doubt we'll see em' earlier than 2Q 2007. Die size seems a little large but doable for AMD, hopfully they can get it down to something a bit more reasonable to improve thier margins.

---

It's only 150mm^2, that's a lot smaller than the dual cores they make now. AMD should get very good yields. You probably looked at the picture and forgot the quad core is 65nm. This makes it really only half the size it looks like in the picture compared to the dual core.

Quote:

---

Originally Posted by Khenglish

It's only 150mm^2, that's a lot smaller than the dual cores they make now.

---

If these pictures are of a 300 mm wafer (and they appear to be), then it's closer to the 283 mm^2 accord99 mentions.

These are pictures of their quad cores. 283mm^2 makes sense

I'm really confused by all these AMD core revisions...

nobody tested anyone yet??

i really would like to see a processor the nock out the conroe!! :P

Quote:

---

Originally Posted by ozzimark

i think he means having one core that clocks worse than the others :D

---

Wouldn't matter since each core has a seperate PLL
thus you can overclock each core to its max

Quote:

---

Originally Posted by nn_step

Wouldn't matter since each core has a seperate PLL
thus you can overclock each core to its max

---

i was under the impression that they all had the same pll, but could adjust multiplier and voltage independantly..

either way, the rough effect is the same :D

Quote:

---

Originally Posted by engenheiro_ce

nobody tested anyone yet??

i really would like to see a processor the nock out the conroe!! :P

---

Earliest you could expect some leaked benches would be Dec. this year or Jan 2007, I'd lean towards the former rather than the latter, despite all the naysayers who said they couldn't pull it off it appears that AMD is following thier schedule and will be on time.

http://www.hkepc.com/bbs/attachments...1GfRYfDBpe.jpg

they say next-gen AMD cpus aka K8L will be 40% faster than todays K8, and thus 20% faster than C2D (Yorkfield is just Kentsfield shrink with SSE4)

Quote:

---

Originally Posted by MAS

[IMG]
they say next-gen AMD cpus aka K8L will be 40% faster than todays K8, and thus 20% faster than C2D (Yorkfield is just Kentsfield shrink with SSE4)

---

At stock speeds.

What about overclocked? We all know conroe overclocks like crazy, and unless these new AMD chips can do the same, I don't see them catching conroe anytime soon :(

Quote:

---

Originally Posted by mr_mordred2095

At stock speeds.

What about overclocked? We all know conroe overclocks like crazy, and unless these new AMD chips can do the same, I don't see them catching conroe anytime soon :(

---

true that!

well for bragging rights that is. Fortunatally most server farms run at stock frequencies.


Obi

in multithreaded apps (rendering mostly) it will spank every conroe even if you OC the later.

tho the hkepc list has some mistakes.

Quote:

---

Originally Posted by MAS

one more pict.
http://www.overclockers.ru/images/ne...17/quad_01.jpg

Kent will have 280 square mm chip surface, K8L - only 150 sq. mm

---

Rev G Quad Core and Rev H Quad Cores are different.

Rev G Quad Core is merely 2 65nm dual core dies on one proc, while Rev. H will be native quadcore i.e. designed from the inside out as a quad core model.

The only match Intel will have to Rev H is Yorkfield. Conroe, penryn, kentsfield will all be architecturally behind the K8L aka Rev H. The advantage to Penryn in 2008 being the ability to have an Octa-Core Core2 model with the same die size as the smithfield (ironic)...how pratical that'll be knowing the yorkfield will be out for a about 8 months by that time)

PLEASE PLAN CAREFULLY FOR UPGRADES WITH THIS INFO... Lots of suckers are gonna be going around shouting "The K8L are here, The K8L are here" when the first quad cores are announced.

Increasing that confusion will be the fact that AMD will introduce some Rev H products with two cores disabled, hence a dual core Rev H.

Perkam

Quote:

---

Originally Posted by perkam

Rev G Quad Core and Rev H Quad Cores are different.

Rev G Quad Core is merely 2 65nm dual core dies while Rev. H will be native quadcore i.e. designed from the inside out as a quad core model.

The only match Intel will have to Rev H is Yorkfield. Conroe, penryn, kentsfield will all be architecturally behind the K8L aka Rev H.

PLEASE PLAN CAREFULLY FOR UPGRADES WITH THIS INFO... Lots of suckers are gonna be going around shouting "The K8L are here, The K8L are here" when the first quad cores are announced.

Increasing that confusion will be the fact that AMD will introduce some Rev H products with two cores disabled, hence a dual core Rev H.

Perkam

---

Perkam,there is NO RevG QC!MCM isn't possible with K8 since two cores can't share one mem. controler...
RevH is the only QC design and the rivals for Kentsfield will be 4x4 systems(at least while there is no K8L in those configurations)

What about coldbug? :D

Quote:

---

Originally Posted by informal

Perkam,there is NO RevG QC!MCM isn't possible with K8 since two cores can't share one mem. controler...
RevH is the only QC design and the rivals for Kentsfield will be 4x4 systems(at least while there is no K8L in those configurations)

---

Quote:

---

Originally Posted by Dailytech

Earlier AMD roadmaps have revealed that quad-core production CPUs would not utilize a native quad-core design until late 2007 or 2008. To put that into perspective AMD demonstrated the first dual-core Opteron samples in August 2004, with the processor tape out in June 2004. The official launch of dual-core Opteron occurred on April 21, 2005. On the same call Meyer announced that that the native quad-core would launch in the middle of 2007 -- suggesting the non-native quad-core Deerhound designs may come earlier than expected or not at all.

---

Something I've been shouting about since March...there ARE non-native quad core designs in AMD's test labs...they just haven't taken a firm decision to launch it so close to the K8L launch as the kentsfield will be unarguably faster, which could cause a long term market stigma of ALL amd quad cores being slower than ALL intel quad cores, thats not to say that the market is stupid...just that the assumption makes it easier to make decisions :p:

Perkam

Quote:

---

Originally Posted by brentpresley

Perkam - did you forget your coffee this morning? Rev G quadcores? Um, not happening.

And with Rev H (K8L) coming in Q2 or Q3 '07, AMD will not have the performance lead for long until it is challenged by Yorkfield (please note I said CHALLENGED, we have no indication yet how K8L and Y-field will compare in IPC and raw clock speed).

---

So you do know how high AMD can clock RevG??
If you don't know,how can you tell AMD can't challenge C2D?
I'm speaking of def. out of the box comparisons,like 3.4GHz Brisbane vs 2.93 C2D ie.We saw that there are 3GHz Windsors coming this Nov.,so Brisbanes that will be Rev G2 or G3(hypothetical numbers) and will incorporate new and improved trensistor mix on 65nm could probably easily hit 3.4GHz(and btw,don't bring those roadmaps showing Windsor based FXes-64 or whatever- in Q3 07 into this discussion,since those are highly inaccurate)

Quote:

---

Originally Posted by perkam

Something I've been shouting about since March...there ARE non-native quad core designs in AMD's test labs...they just haven't taken a firm decision to launch it so close to the K8L launch as the kentsfield will be unarguably faster, which could cause a long term market stigma of ALL amd quad cores being slower than ALL intel quad cores, thats not to say that the market is stupid...just that the assumption makes it easier to make decisions :p:

Perkam

---

Perkam,as I read that,it says that they(dailytech) seems to think he suggested that,not that he actually said:" There are non-native QC on the roadmap in 07".Dailytech has a vivid imagination :p:
Read that again,maybe you will see it too :)

Quote:

---

Originally Posted by brentpresley

Perkam - did you forget your coffee this morning? Rev G quadcores? Um, not happening.

And with Rev H (K8L) coming in Q2 or Q3 '07, AMD will not have the performance lead for long until it is challenged by Yorkfield (please note I said CHALLENGED, we have no indication yet how K8L and Y-field will compare in IPC and raw clock speed).

---

AMD and Intel have both confirmed that their nex gen native QC designs are launching Q3 '07...

All this talk about AMD "will be behind" is crap...it'll on par with the year...Conroe allowed Intel's crown to remain for 1 year, while AMD's lasted 3.

Perkam

Quote:

---

Originally Posted by brentpresley

The contact I mentioned a few posts above (DOD) has hinted to me that the 65nm ES chips they have do not OC well at all. Although he won't come out and just say it or put it in writing b/c of the NDAs.

NOW - I don't know how early on the chips he has are, or how much AMD has improved things since then. So keep that in mind.

But to expect that a dumb shrink (Rev G) of K8 is going to overtake Core 2 (20% faster IPC) is just wishful thinking.

Intel has had a YEAR to perfect their 65nm SHIPPING products. AMD is still perfecting 65nm FOR RELEASE. Simply put, Intel has had MUCH longer to work with the process technology and improve both the yields and the speed their chips can hit.

Additionally, it looks like we are seeing further improvements in Intel's 65nm process tech. Early release Core 2 chips had a lot of difficulty getting above 3.5GHz (unless you had a REALLY cherry-picked CPU). Now it seems that hitting 3.7-3.8GHz on decent air cooling is not that hard, if you have a motherboard that is up to the task of the high FSBs.

For Reg G(2, 3, whatever) to MATCH the performance of a 3.5GHz Core 2, AMD would have to get the clock speed up to 3.5GHz * 1.2 = 4.2GHz, assuming a 20% IPC lead for Core 2. I won't say it is impossible, but in a SHIPPING product it is going to take QUITE A WHILE.

No AMD's answer to the Core 2 performance crown is Rev. H. Period.

---

The performance in games is much closer than you would think Brent.

Perkam

Quote:

---

Originally Posted by brentpresley

The contact I mentioned a few posts above (DOD) has hinted to me that the 65nm ES chips they have do not OC well at all. Although he won't come out and just say it or put it in writing b/c of the NDAs.

NOW - I don't know how early on the chips he has are, or how much AMD has improved things since then. So keep that in mind.

But to expect that a dumb shrink (Rev G) of K8 is going to overtake Core 2 (20% faster IPC) is just wishful thinking.

Intel has had a YEAR to perfect their 65nm SHIPPING products. AMD is still perfecting 65nm FOR RELEASE. Simply put, Intel has had MUCH longer to work with the process technology and improve both the yields and the speed their chips can hit.

Additionally, it looks like we are seeing further improvements in Intel's 65nm process tech. Early release Core 2 chips had a lot of difficulty getting above 3.5GHz (unless you had a REALLY cherry-picked CPU). Now it seems that hitting 3.7-3.8GHz on decent air cooling is not that hard, if you have a motherboard that is up to the task of the high FSBs.

For Reg G(2, 3, whatever) to MATCH the performance of a 3.5GHz Core 2, AMD would have to get the clock speed up to 3.5GHz * 1.2 = 4.2GHz, assuming a 20% IPC lead for Core 2. I won't say it is impossible, but in a SHIPPING product it is going to take QUITE A WHILE.

No AMD's answer to the Core 2 performance crown is Rev. H. Period.

---

Well we all know tha AMD ES usually suck when compared to shipping cpus.This has been the regular case in the past.
So,i said if we compare out of the box products that will come Q1 07 and IF AMD can ramp up clock on Brisbanes,3.4GHz would be enough for C2D to match it in 90% of the cases.
Remember one thing: usual everyday peple don't know or care about OCing.They can read some PCMagazin or whatever and if they see the chips are on par and the prices are on par also,majority of folks will look at TOTAL cost of the platform and decide what to buy

this should be interesting :D

Quote:

---

Originally Posted by brentpresley

Possible, but my gut tells me that if Intel got even a HINT that AMD was going to try to shoot for the performance crown I bet we would "magically" see some 3.2 - 3.4GHz Core 2 Extreme editions pop up FS. I wouldn't be surprised if Intel has binned a few dozen thousand chips that clock well just for a situation like this.

Keep in mind, we are not dealing with the same Intel from the past 10-15 years that was pretty comfortable sitting on it's laurels and letting AMD have x% of the market.

Netburst was a mistake. A mistake that cost Intel the performance crown for about 3 years. Intel is PISSED about the loss of face AND the loss of marketshare. I don't think we will see them make a MAJOR screwup like Netburst again for a LONG LONG TIME.

Have no doubts, we are witnessing the first salvos in a war of attrition. Intel will take the lead, and them AMD will take it back. And Vice Versa for YEARS to come.

The ONLY winners are going to be us, the users.

Welcome to the NEW ERA of dirt cheap super-chips.

(now if we could just get this kind of war in the GPU arena we would be set).

---

Very true...we'll have Super computer FPU performance in 2008...on a $500 PC.

Definitely is interesting :)

Perkam

Quote:

---

Originally Posted by brentpresley

Have no doubts, we are witnessing the first salvos in a war of attrition. Intel will take the lead, and them AMD will take it back. And Vice Versa for YEARS to come.

The ONLY winners are going to be us, the users.

Welcome to the NEW ERA of dirt cheap super-chips.

(now if we could just get this kind of war in the GPU arena we would be set).

---

Well said, it's a beautiful time for us end users, and pc enthusiasts. :p:

Quote:

---

Originally Posted by perkam

AMD and Intel have both confirmed that their nex gen native QC designs are launching Q3 '07...

All this talk about AMD "will be behind" is crap...it'll on par with the year...Conroe allowed Intel's crown to remain for 1 year, while AMD's lasted 3.

Perkam

---

1. Intel has the lead over AMD in x86 CPU for ages.
2. Netburst was a tragic mistake that Intel used for the past 6 years.
3. Intel's product development cycle is much shorter now with their new strategy. There's only a slim chance for AMD to keep up.
4. The best part of AMD's business will now be its GPU, aside from its digital TV and cellphone coprocessors.

What do we know today?
1. Server K8l Deerhound Opteron will be announced in Q2 2007
2. FX socket F+ (with HT 3.0 support) for 4x4 (in fact it's the same Opteron, remarked) and X4 am2+ (also with HT 3.0 support) - in Q3 2007
3. Dell will buy a great quantity of AMD procs :)
4. In several months AMD will show us the system with K8L and proove its high performance. Maybe this will be octal-core 4x4 :)
5. 65nm AMD procs can operate at clocks not less than C2D B2, B3
6. There is no quad-core K8 in AMD plans at all, quad-core is only K8L

brentpresley, look at overclock abilities of 90nm Dothan and K8 - frequency limit (temperature 20C) is just the same
why OC limit for 65nm short-stage intel and amd cpus must differ? (sorry for bad engl. :) )

Quote:

---

Originally Posted by vitaminc

3. Intel's product development cycle is much shorter now with their new strategy. There's only a slim chance for AMD to keep up.

---

I wouldn't bet the farm on that ;)

http://www.xtremesystems.org/forums/...d.php?t=118091

Quote:

---

Originally Posted by MAS

brentpresley, look at overclock abilities of 90nm Dothan and K8 - frequency limit (temperature 20C) is just the same
why OC limit for 65nm short-stage intel and amd cpus must differ? (sorry for bad engl. :) )

---

So it's speculation, not an fact.

The first few batch of 65nm AM2 CPU probably won't clock so well. It's a new process node for AMD afterall.

Quote:

---

Originally Posted by LowRun

I wouldn't bet the farm on that ;)

http://www.xtremesystems.org/forums/...d.php?t=118091

---

GP GPU has nothing to do with CPU. Sure GPUs (ATI > nVidia) has :banana::banana::banana::banana:loads of FLOPs, but all of the GPU will just choke so hard when running x86 style instructions in x86 CPU thermal envelopes.

Hopefully we could use a 200W PSU for a CrossFire or SLi high end gaming rig someday.

Quote:

---

Originally Posted by vitaminc

So it's speculation, not an fact.

The first few batch of 65nm AM2 CPU probably won't clock so well. It's a new process node for AMD afterall.

---

True, however the first few batches are about 1 Month of Time. After that you have to deal with AMD's Rapidly evolving process, realtime process changes combined with a group of highly skilled and motivated Individuals working as a team. At the same time for the first 65m chips (provided that are K8) From AMD would clock comparative to Yohan (Conroe's father) and shortly after it'll start to clock faster and faster

Quote:

---

Originally Posted by nn_step

True, however the first few batches are about 1 Month of Time. After that you have to deal with AMD's Rapidly evolving process, realtime process changes combined with a group of highly skilled and motivated Individuals working as a team. At the same time for the first 65m chips (provided that are K8) From AMD would clock comparative to Yohan (Conroe's father) and shortly after it'll start to clock faster and faster

---

We could debate AMD's APM vs. Intel's Copy Exact strategy all day long, but the fact is that yields defect density will only go down with time.

You can never state the overclockability of a chip as a fact before you have them on hands with actual data. Everything now is speculations under NDA smoke screens.

Quote:

---

AMD's chips made EARLY on in a process shrink (90nm, 65nm, etc.) don't overclock very well. They continue to improve the process and release faster and faster chips. It just takes some time, but they eventually get VERY good with working with a particular process level.

---

some times i think that amd does rthis stuff on purpose , just look at the mhz that the 4x4 cpu's are going to be , amd has no other cpu running that fast

Quote:

---

(they have the luxury of having an ENTIRE EXPERIMENTAL FAB - D1D in Oregon),

---

i read some where that amd has , as you would say it , a experimental lab at fab 30 and fab 36

Quote:

---

Originally Posted by The Ghost

some times i think that amd does rthis stuff on purpose , just look at the mhz that the 4x4 cpu's are going to be , amd has no other cpu running that fast


i read some where that amd has , as you would say it , a experimental lab at fab 30 and fab 36

---

Actually ALL AMD fabs include the ability to experiment and improve process performance and yeilds

Quote:

---

Originally Posted by The Ghost

i read some where that amd has , as you would say it , a experimental lab at fab 30 and fab 36

---

They don't. They jointly develope with IBM. They do have a "test" line for future processes (not a whole dedicated fab like Intel).

Quote:

---

Originally Posted by nn_step

Actually ALL AMD fabs include the ability to experiment and improve process performance and yeilds

---

The improve process performance/yields are result of APM (software fab control and other magics). Has nothing to do with their process development for 45nm or 32nm.

Quote:

---

Originally Posted by vitaminc

GP GPU has nothing to do with CPU. Sure GPUs (ATI > nVidia) has :banana::banana::banana::banana:loads of FLOPs, but all of the GPU will just choke so hard when running x86 style instructions in x86 CPU thermal envelopes.

Hopefully we could use a 200W PSU for a CrossFire or SLi high end gaming rig someday.

---

Looks like you didn't even read the article :rolleyes:

Quote:

---

Originally Posted by LowRun

Looks like you didn't even read the article :rolleyes:

---

skim through it. nothing interesting that i havent heard of, except it does not address a lot of important concerns such as power envelope, process integration, etc.

Quote:

---

Originally Posted by perkam

The performance in games is much closer than you would think Brent.

Perkam

---

So will the performance in wordpad. Just goes to show that games that are out now do not utilize CPU's well. Though I think they will in the future, what with physics simulation and all.

Also Perkam,i totaly forgot about this : Did you ever thought that S.Meyer of AMD might be talking about 4x4 as non-native quad core solution?I did,and it goes well in line with what H.Richards said at Digitimes interviews when asked about that.
So i still am 99.9% sure there never will be MCM K8 based non native QC.As i said it previously,how is it possible that both chips share one mem. controler(be it on first or second core) and still use the same socket?OTOH,I can see that two sockets can make non native=almost MCM like solution= and that is what AMD publicly stated as their plan to combat Kents(which may fair well against since there could be some serious cache trashing -during heavy multitasking- in Kentsfield's shared cache structure and FSB bottlenecking,no matter if it is not full 8MB of shared L2,but 2x4MB)

Slightly OT: As for Daily tech credibility lately,look at this one :http://images.dailytech.com/nimage/2...07_roadmap.png
Tell me they are serious,since they have been wrong with their "roadmaps" (custom made ? :p: ) when they spoke about "roadmaps" showing AMD only getting to X2 5600+ in Q3'07...

Quote:

---

Originally Posted by vitaminc

skim through it. nothing interesting that i havent heard of, except it does not address a lot of important concerns such as power envelope, process integration, etc.

---

Must be your way of saying you've only read the title otherwise you wouldn't say something like

Quote:

---

GP GPU has nothing to do with CPU. Sure GPUs (ATI > nVidia) has loads of FLOPs, but all of the GPU will just choke so hard when running x86 style instructions in x86 CPU thermal envelopes.

---

Even without reading the article i can't see why one would want to run x86 style instructions on a GP-GPU that sits on a die with 4 x86 cores by it's side.
They won't put it there for you to run x86 instructions :slap:

Quote:

---

Originally Posted by informal

Also Perkam,i totaly forgot about this : Did you ever thought that S.Meyer of AMD might be talking about 4x4 as non-native quad core solution?I did,and it goes well in line with what H.Richards said at Digitimes interviews when asked about that.
So i still am 99.9% sure there never will be MCM K8 based non native QC.As i said it previously,how is it possible that both chips share one mem. controler(be it on first or second core) and still use the same socket?OTOH,I can see that two sockets can make non native=almost MCM like solution= and that is what AMD publicly stated as their plan to combat Kents(which may fair well against since there could be some serious cache trashing -during heavy multitasking- in Kentsfield's shared cache structure and FSB bottlenecking,no matter if it is not full 8MB of shared L2,but 2x4MB)

---

According to HKEPC (Hong Kong based site), Yorksfield will be Intel's 45nm single die quad core with Conroe core in 2H07.

FSB problem still presists, but it's not as bad as people made it out to be.

Quote:

---

Originally Posted by informal

Slightly OT: As for Daily tech credibility lately,look at this one :http://images.dailytech.com/nimage/2...07_roadmap.png
Tell me they are serious,since they have been wrong with their "roadmaps" (custom made ? :p: ) when they spoke about "roadmaps" showing AMD only getting to X2 5600+ in Q3'07...

---

Well, if you think about it, AMD has always been focusing on architecture improvement since the K6 days, thus the clocking of X2/X4 are not that big of a deal IMO. They could always do it Intel P4 style and clock it really high, but that's not "green" computing. :p

Quote:

---

Originally Posted by LowRun

Must be your way of saying you've only read the title otherwise you wouldn't say something like



Even without reading the article i can't see why one would want to run x86 style instructions on a GP-GPU that sits on a die with 4 x86 cores by it's side.
They won't put it there for you to run x86 instructions :slap:

---

So they only taught you reading but not thinking in school?

"X86 style" instructions meaning codes that's not highly parallel like graphics.

1. You don't think that all of the current HPC softwares will be recompiled away from the current x86/SPARC/PPC codes and translate into DX9/DX10 based GPU code, do you?
2. Running a couple X1950XTX 512M cards with 200W power dissipation each under load just all of the sudden makes a whole lot of sense, especially with AMD's compaign of green computing.

Brent is right about 65nm

Even AMD says the same, they silently add new tech bits to their sauce. Early on in a shrink the process is almost exactly the same, and basically results in only a power and die size shrink. Further out when they begin adding the new tech is when the clocks start to ramp consistently. (i did have an early winchester that clocked to 2.77ghz which was great for the time, there will always be a few exceptions)

Quote:

---

Originally Posted by vitaminc

According to HKEPC (Hong Kong based site), Yorksfield will be Intel's 45nm single die quad core with Conroe core in 2H07.

FSB problem still presists, but it's not as bad as people made it out to be.

---

I was refering to Novembre release of Kentsfield and the AMD's response to that(in my opinion,it's not MCM K8 like hybrid-gotta love that term:p:).
As for Yorksfield,i think it will be somewhat better than Kents ,BUT if it has no IPC improvements,i don't know how well it will fare against K8L,especially in HPC segment(in all honesty I believe that server variant of K8L will be better,and by a wider margin)

Quote:

---

Originally Posted by vitaminc

Well, if you think about it, AMD has always been focusing on architecture improvement since the K6 days, thus the clocking of X2/X4 are not that big of a deal IMO. They could always do it Intel P4 style and clock it really high, but that's not "green" computing. :p

---

I was speaking of Brisbanes arrival dates in that DailyTech "roadmap".It's said that they will arrive in Q1 07,but other sources (HKPEC for instance) are saying Decembre.As for clock speeds on that slide,they are in line with what we know by now,my only objection was the fact that dailytech wasn't upto the mark,at least when roadmaps for AMD are in question

http://www.aceshardware.com/forums/r...7742&forumid=1

Quote:

---

They are flat out wrong. 65nm is set, has been set for a long time, and knowing the exact date, I can say they are wrong. Trust them at your own peril.

-Charlie

---

Charlie is one of the most informed in IT industry(by my merits at least),and I trust him on this.When he says "has been set for a long time",it seems to me that Brisbanes are over their initial problems and are set to go into retail in Dec(this year :) )

Quote:

---

Originally Posted by vitaminc

So they only taught you reading but not thinking in school?

"X86 style" instructions meaning codes that's not highly parallel like graphics.

1. You don't think that all of the current HPC softwares will be recompiled away from the current x86/SPARC/PPC codes and translate into DX9/DX10 based GPU code, do you?
2. Running a couple X1950XTX 512M cards with 200W power dissipation each under load just all of the sudden makes a whole lot of sense, especially with AMD's compaign of green computing.

---

1. Obviously if an app is to take advantage of a GP-GPU, is has to be coded specificaly :rolleyes:

2. I don't even see what kind of point you're trying to make there.

Quote:

---

Originally Posted by LowRun

1. Obviously if an app is to take advantage of a GP-GPU, is has to be coded specificaly :rolleyes:

2. I don't even see what kind of point you're trying to make there.

---

1. Do you recollect the debate between RISC vs CISC and how road of domination by x86 instruction sets.

2. GPU has a lot of FLOPS, but it doesnt mean that they are cost effective solution. One single X1950XTX consumes as much as 3 Socket F Opterons (68W HE version TDP) or as much as 4 C2Q Kentsfields (50W ULV version TDP). HVAC is a huge concern with GPUs, and AMD's success with Opteron is mostly due to its high power efficiency.

Quote:

---

Originally Posted by vitaminc

2. GPU has a lot of FLOPS, but it doesnt mean that they are cost effective solution. One single X1950XTX consumes as much as 3 Socket F Opterons (68W HE version TDP) or as much as 4 C2Q Kentsfields (50W ULV version TDP). HVAC is a huge concern with GPUs, and AMD's success with Opteron is mostly due to its high power efficiency.

---

might be 3x the power, but it's WAAAYYY more than 3x the FP throughput ;)

Quote:

---

Originally Posted by vitaminc

1. Do you recollect the debate between RISC vs CISC and how road of domination by x86 instruction sets.

2. GPU has a lot of FLOPS, but it doesnt mean that they are cost effective solution. One single X1950XTX consumes as much as 3 Socket F Opterons (68W HE version TDP) or as much as 4 C2Q Kentsfields (50W ULV version TDP). HVAC is a huge concern with GPUs, and AMD's success with Opteron is mostly due to its high power efficiency.

---

1. Don't know if it's my english reaching it's limits but i don't get what you mean there. I'm just refering to what Stanford did with folding@home for example, they coded a version specificaly to take advantage of the GP-GPU.

2. You keep on bringing that x1950xtx to the debate, did i said AMD was going to integrate one on their CPUs? Nope. In fact noone did. I don't know how they would do it but i'm confident they would do very well and as stated in the article that would put them in a very good position.

Quote:

---

Originally Posted by LowRun

1. Don't know if it's my english reaching it's limits but i don't get what you mean there. I'm just refering to what Stanford did with folding@home for example, they coded a version specificaly to take advantage of the GP-GPU.

---

On top of DirectX 9 and ATI specific driver.

Quote:

---

Originally Posted by LowRun

2. You keep on bringing that x1950xtx to the debate, did i said AMD was going to integrate one on their CPUs? Nope. In fact noone did. I don't know how they would do it but i'm confident they would do very well and as stated in the article that would put them in a very good position.

---

What does CPU/GPU integration has to do with power envelope? Having the capability in integrating GPU/NB/MC into GPU is one thing; doing the integration within a tight power envelope is another. Who care about where each individual component reside as long as it met power requirement and system level performace is competitive. I have faith in both AMD and Intel in the integration process, but to satisfy the power envelop requirement is another. Besides, Intel is the expert in IGP with approximately 50% marketshare.

All integrated graphics from nVidia and ATI have are using previous generation technologies, so I will not live my hope up for a massive integrated 350W TDP K8L + X1950.

well no one says they'd use an x1950 type of core. Seems to me that any kind of core that specializes in graphical/physics/FP operations would make a tremendous boost. Perhaps x16xx series? Even x13xx series? There's more than one option.

Ryan

so many info in this post.. but for me.. I'll wait for it.. :D

Quote:

---

Originally Posted by vitaminc

On top of DirectX 9 and ATI specific driver.

---

That make my point even more

Quote:

---

What does CPU/GPU integration has to do with power envelope? Having the capability in integrating GPU/NB/MC into GPU is one thing; doing the integration within a tight power envelope is another. Who care about where each individual component reside as long as it met power requirement and system level performace is competitive. I have faith in both AMD and Intel in the integration process, but to satisfy the power envelop requirement is another. Besides, Intel is the expert in IGP with approximately 50% marketshare.

All integrated graphics from nVidia and ATI have are using previous generation technologies, so I will not live my hope up for a massive integrated 350W TDP K8L + X1950.

---

Once again, you're the only one thinking x1950 about AMD's GP-GPU integration into their CPUs, don't know how you got there. I believe they will do fine regarding power envelope with a solution specificaly designed for the task. As for Intel being the expert in IGP, having 50% of the market share doesn't make them experts IMO, this is due to many reasons that have nothing to do with performance or efficiency, in fact their IGPs are utterly crap and they are lagging behind badly in that field wich could put them in a bad position when AMD will bring their GPU on CPU.

With all due respect, I think Vitaminc is confusing the recently released F@H client for the X1900 cards with utilizing a GPU type processor as a coprocessor for the CPU.

I think ti's perfectly clear that AMD will use these types of coprocessors, in fact many different types, anything that will go on the HT3.0 bus. It would be a very logic first step for them to start with using Ati's GPU tech.

Ofcourse there are technical difficulties that I cant even comprehend, but that shouldt stop the engineers from finding solutions to tap into this extra power. Else we would still be using 286 CPU's without a seperate math coprocessor. (I believe this was added in the 386 DX, correct me if Im wrong.)

Low Run, you are dead wrong on this. GPU on CPU is a wet dream for AMD at the momemnt for 3 reasons:
1) power envelope for integrated graphics are simply way too high.
2) graphics engine evolves faster than moore's law and requires frequent updates.
3) a good one year lag in manufacturing technologies.

I am sure all the IT departments and HPC institutions that you know just loves to apply new graphics drivers (CPU if you believe GPU on CPU story) every couple weeks, new DX (or whatever API) every year, and patch your HPC code as frequent as the combination of the two.

A tailor made x86 threaded application on x86 processors (Opteron or C2D) should be able to chuck out nearly as many FLOPs as a GPU, in the same power envelope. All those x86-64 and SSE1-4 extentions are implimented for this.

Vodka,
The match coprocessor pre-pentium Intel processors are x87 series. 8087 for 8086, 80287 for 80286, 80387 for 80386, 80487 for 80486. Their main purposes are to add FLOPs to the system. No one cares if the coprocessor resides on FSB, PCI-express x16, HT 1.0/3.0, or CSI, as long as its profitable for the co-processor company to design and manufacturer and the power/performance envelope is reasonable.

Guys let us stick to the thread title and not wander off to 2008 or 2009 with this GPU on CPU stuff.
Topic is called "AMD K8L ES coming in Decembre":rolleyes:
As far as we all know it has no GPU onboard so we shouldn't go into this stuff ,at least not in this thread.Start a new one and battle in there :)

Quote:

---

Originally Posted by informal

Guys let us stick to the thread title and not wander off to 2008 or 2009 with this GPU on CPU stuff.
Topic is called "AMD K8L ES coming in Decembre":rolleyes:
As far as we all know it has no GPU onboard so we shouldn't go into this stuff ,at least not in this thread.Start a new one and battle in there :)

---

I will second that :clap:

Quote:

---

Low Run, you are dead wrong on this. GPU on CPU is a wet dream for AMD at the momemnt for 3 reasons:
1) power envelope for integrated graphics are simply way too high.
2) graphics engine evolves faster than moore's law and requires frequent updates.
3) a good one year lag in manufacturing technologies.

---

this will happen at 45nm cpu's in 2008 for notebooks

Quote:

---

A tailor made x86 threaded application on x86 processors (Opteron or C2D) should be able to chuck out nearly as many FLOPs as a GPU, in the same power envelope. All those x86-64 and SSE1-4 extentions are implimented for this

---

.
this is not true or software writers would have done this already , amd has many partners that are developing co-processors to do this , ati is one of these partners that are developing co-processors

by the way intel has said something about putting a gpu on a cpu

http://www.gpgpu.org/

Quote:

---

Originally Posted by The Ghost

this will happen at 45nm cpu's in 2008 for notebooks


.
this is not true or software writers would have done this already , amd has many partners that are developing co-processors to do this , ati is one of these partners that are developing co-processors

by the way intel has said something about putting a gpu on a cpu

http://www.gpgpu.org/

---

1. I believe AMD has not yet revealed any roadmaps for its notebook processors beyond 2007. Correct me if I am wrong though.
2. AMD's notebook processors is using a different design than its desktop/server procesors, as disclosed in June 2006 Tech Analyst Day.
3. Most softwares aren't very capable of using multi-core CPU yet, and has been a concern for both Intel and AMD.
4. As I said before, co-processors will be better, but requires very specific software support.

Quote:

---

Originally Posted by vitaminc

1. I believe AMD has not yet revealed any roadmaps for its notebook processors beyond 2007. Correct me if I am wrong though.
2. AMD's notebook processors is using a different design than its desktop/server procesors, as disclosed in June 2006 Tech Analyst Day.
3. Most softwares aren't very capable of using multi-core CPU yet, and has been a concern for both Intel and AMD.
4. As I said before, co-processors will be better, but requires very specific software support.

---

vitaminc,i appreciate your effort to bring the info about CPU-GPU merging and notebook usage of those designs,but pls read : http://www.xtremesystems.org/forums/...3&postcount=95

Cheers

Quote:

---

1. I believe AMD has not yet revealed any roadmaps for its notebook processors beyond 2007. Correct me if I am wrong though.

---

phil hester said so in his last interview with zdnet
http://insight.zdnet.co.uk/hardware/...83795-2,00.htm
Are there changes that you're planning to make to the core for the mobile space?
One of the areas we need to work on as a company is the mobile space. And that's where the biggest win comes, from being able to integrate the graphics.


Integration in the microprocessor itself or integration in the chipset?
Integration of the CPU and the GPU. Assuming the transaction closes on time, we would target a merged design in the 45nanometre time frame.


Which is 2008?
Yeah. Another thing happening in the graphics space is that there's more and more programmability. It used to be that it was just polygon rendering. That's what graphics was, but now developers are doing so much programming.


The next generation of gaming is really making things more dynamic. It's not making the surface look realistic, but making it behave realistically. We've crossed the point where the GPU can do real programs of a significant size.


It may seem like 2008 is a long way away, but that's a major design cycle. ATI also has very good business, in the handset and set-top box DTV area.

Quote:

---

4. As I said before, co-processors will be better, but requires very specific software support.

---

co-processors work with the software that is out there already , ati could be one of those co-processors

Quote:

---

Originally Posted by vitaminc

1. Intel has the lead over AMD in x86 CPU for ages.

---

Meh, the whole ages comment seems overblown unless you're referring to ages ago. They've been trading the lead since the K7 intro. with things working out more in favor of AMD as far as performance goes...

Quote:

---

Originally Posted by vitaminc

2. Netburst was a tragic mistake that Intel used for the past 6 years.

---

Yea, only good thing about it was it gave them the lead again over the at the time aging AthlonXP's til' A64 came out, still sold like hotcakes though. :/

Quote:

---

Originally Posted by vitaminc

3. Intel's product development cycle is much shorter now with their new strategy. There's only a slim chance for AMD to keep up.

---

Isn't this assuming perfect execution on the design and process side? Haven't they already had some minor delays for the intro of thier 45nm process? You cut Intel this slack but won't do the same for AMD? What gives?

FWIW I'd say they're both going to be trading the lead back and forth a whole lot the next 4-6yr. or so assuming that niether makes any major screw ups.

Quote:

---

Originally Posted by vitaminc

4. The best part of AMD's business will now be its GPU, aside from its digital TV and cellphone coprocessors.

---

This seems unreasonable (completely disregards any and all of AMD's future CPUs) and way to specualtive (we have no idea how AMD's on die/package GPU strategy will pan out or what they plan on doing exactly for high end GPU products...). You're also completely disregarding any attempt at them working on chipsets too...

WTH happened vitc? I know you got a AMD system spec in your sig and all but are you getting paid by Intel to say this stuff or you just feeling pessimisstic or what?

Ghost,

GPU on CPU in general if you quote it out of context, but GPU on mobile CPU if you read the paragraph above and below. I would still argue vision and concrete roadmaps are two different things.

Mesyn,

If you look at the history of both AMD and Intel, Intel has the lead in significantly portion of the time in processor designs and has always been the leader in process/manufacturing technologies (unless you want to argue about apm vs. copy exact).

Not sure where you get the words on Intel's 45nm delays. They are finishing up 1 fab and building 2 extra ones to copy exact, as per IDF. Care to elaborate?

ATI's digital TV and cellphone multimedia coprocessors are the market leaders in their respective market. And those 2 markets are growing helluva fast. Desktop PC are in a decline as people switching to notebooks, thus AMD's attempt to get its own platform to battle Centrino.
I believe that big screen LCD/plasma TV will grow faster than PC for the next few years and people will always change cellphone faster than they change computers, thus my speculation that AMD's crown jewel will be in those 2 markets (if they don't mess up). :p
Unless you believe that PC growth will be faster than big screen LCD/plasma TV, or people changing

I could diss Intel all day long (like its bureacracy, FSBrenza, IGP), but that's not related to Altair. :p

Quote:

---

GPU on CPU in general if you quote it out of context, but GPU on mobile CPU if you read the paragraph above and below. I would still argue vision and concrete roadmaps are two different things.

---

roadmaps and visions are the same thing , there is no such thing as a concrete roadmap , we have seen things that have been on amd roadmaps that never came too , and we have seen things on intels roadmaps that never came too

also if you have seen some roadmaps , and have listened to members from amd , you would have noticed that they are also going to use mobile cpu's on low end desktops , so yes we can end up with desktop with intergrated gpu in to a cpu , companies like sell would jump all over something like this , where they are not worried about high end graphics

Quote:

---

Originally Posted by The Ghost

roadmaps and visions are the same thing , there is no such thing as a concrete roadmap , we have seen things that have been on amd roadmaps that never came too , and we have seen things on intels roadmaps that never came too

also if you have seen some roadmaps , and have listened to members from amd , you would have noticed that they are also going to use mobile cpu's on low end desktops , so yes we can end up with desktop with intergrated gpu in to a cpu , companies like sell would jump all over something like this , where they are not worried about high end graphics

---

They are different things. Vision is what the executives/marketing people envisioning their future products, such as 10GHz CPU hot wet dream by Intel.

Roadmaps is what companies release to their customers as a product release schedule promise so their customers can better manage inventories and software venders can anticipate the hardware changes.

Mobile on desktop has nothing to do with GPU on CPU. Apple/Shuttle and various other venders already have mobile on desktop PC out on market and those PC tends to use IGP instead of discrete graphics.

There are more engineering concerns in implimenting GPU on CPU for mobile. TDP, die size, pin count, and fan out are the most obvious ones.

Quote:

---

Originally Posted by brentpresley

That is because they are GPU bound. If we are talking RAW CPU performance here, C2D has a 20% edge over K8.

It's not Intel and AMD's fault the GPU companies can't keep up w/ the need for higher pixel fill rates.

---

I dont think its the GPU market either.. its the software companies. They need to come out with software that will utilize an X1950XTX, or 7950GX2.... From the looks of the new 8900GTX Nvidia's card will be about 2x's faster than the 7950GX2... I highly doubt any game will stress that.

~Mike

Quote:

---

Originally Posted by arisythila

I dont think its the GPU market either.. its the software companies. They need to come out with software that will utilize an X1950XTX, or 7950GX2.... From the looks of the new 8900GTX Nvidia's card will be about 2x's faster than the 7950GX2... I highly doubt any game will stress that.

~Mike

---

At the same time microsoft will need better optimized DX9/10 compilers, Nvidia and ATI need better drivers, and game software companies need to multithread certain loads to CPU.

Quote:

---

Roadmaps is what companies release to their customers as a product release schedule promise so their customers can better manage inventories and software vendors can anticipate the hardware changes.

---

do you remember the amd mustang ? wasn't it on a roadmap ??

AMD's Mustang processor was supposed to have support for up to 4MB of L2 cache on the chip die with tweaks to the Athlon Thunderbird core to allow for the addition of more L2 cache on chip, possibly with more pipeline stages, enabling higher clock speeds.
http://www.geek.com/procspec/amd/k7mustang.htm

now do i really need to go down the list of Intel cpus that was on the roadmap and was canceled ??

road maps are company visions , and neither is set in concrete

i know the difference between wet dreams and roadmaps

Quote:

---

Mobile on desktop has nothing to do with GPU on CPU. Apple/Shuttle and various other vendors already have mobile on desktop PC out on market and those PC tends to use IGP instead of discrete graphics.

---

low end desktops are going to use mobile cpu , so yes it has something to do with gpu on CPU , that is a fact , even Intel is trying to do the same thing

i already provided the link that Phil Hester said it would happen , there have been others before him that have said the same thing

Quote:

---

Originally Posted by vitaminc

Mesyn,

If you look at the history of both AMD and Intel, Intel has the lead in significantly portion of the time in processor designs and has always been the leader in process/manufacturing technologies (unless you want to argue about apm vs. copy exact).

---

I was referring to performance, but Intel usually has the process lead.

Quote:

---

Originally Posted by vitaminc

Not sure where you get the words on Intel's 45nm delays. They are finishing up 1 fab and building 2 extra ones to copy exact, as per IDF. Care to elaborate?

---

From this article: http://www.eetimes.com/news/semi/rss...leID=192501516

Perhaps something else has changed that they don't know about because Intel roadmaps show QC 45nm chips available by Q3 2007, though they quote Intel as thier source for thier info...

Quote:

---

Originally Posted by vitaminc

ATI's digital TV and cellphone multimedia coprocessors are the market leaders in their respective market. And those 2 markets are growing helluva fast.

---

What? While growing fast aren't these markets niche as all hell? Hasn't it already been shown that most people don't really play games and stuff on thier cell phones, its just a gimmick? You can't run a 300mm 65nm or for that matter 45nm fab off of profits from this market alone, or for that matter even keep one busy.

Quote:

---

Originally Posted by vitaminc

Desktop PC are in a decline as people switching to notebooks, thus AMD's attempt to get its own platform to battle Centrino.

---

They're still a significant chunk of business though, and they aren't going away any time soon either, you don't see Intel cancelling its desktop chips do you? AMD's efforts to get a competitive platform as a alternative to Centrino is an effort to improve profit margins, not a make or break issue.

Quote:

---

Originally Posted by vitaminc

I believe that big screen LCD/plasma TV will grow faster than PC for the next few years and people will always change cellphone faster than they change computers, thus my speculation that AMD's crown jewel will be in those 2 markets (if they don't mess up). :p
Unless you believe that PC growth will be faster than big screen LCD/plasma TV, or people changing

---

While cellphones are ubiquitos and have a relatively high turn over rate you can't justify the costs of running a high end fab with production for them, I don't see digital TV growth growing massivly either. In fact I think you're gonna see a decline in general sales/profits across the boards for the next few years, the US is heading for a major recession though I don't think you'll see signs of it til' about Q2/3 2007.

Quote:

---

Originally Posted by vitaminc

I could diss Intel all day long (like its bureacracy, FSBrenza, IGP), but that's not related to Altair. :p

---

Well neither is alot of the other stuff you're talking about, you seem to be more interested in how AMD will do financially than about CPU's in general.